The website.dtd, xalan, and other matters
I have had success (woo-hoo) formatting the sample website.xml file that is in
Norm's website.dtd distribution, but with xt, not with xalan_j. Xalan works
with some of its simple examples, but here's what it produces with the
website.xml sample:
--- snip ---
# java org.apache.xalan.xslt.Process -in website.xml -xsl
../xsl/website.xsl -HTML
========= Parsing
file:/home/ruptured-duck/docbook-website-dtd/example/../xsl/website.xsl
========== Parse of
file:/home/ruptured-duck/docbook-website-dtd/example/../xsl/website.xsl
took 17691 milliseconds
========= Parsing
file:/home/ruptured-duck/docbook-website-dtd/example/website.xml
========== http://docbook.org/xml/4.1.2/docbookx.dtd; Line 66; Column 9
XSL Error: Could not parse
file:/home/ruptured-duck/docbook-website-dtd/example/website.xml
document! XSLT: Recursive reference "%dbnotn;". (Reference path:
(top-level)-%docbook;-%dbhier;-%dbnotn;-%dbnotn;) Xalan: was not
successful.
--- snip ---
I mention this both to follow up on recent posts by myself, and because I note
an ITP for xalan has been mentioned on debian-devel. Ergo, here is my little
heads-up wrt xalan (maybe it's common knowledge): that I was successful in
short order with xt suggests to me that there's a shortfall in xalan's ability
to deal with DocBook. Indeed, according to some recent correspondence on
DOCBOOK-APPS, xalan right now is too buggy to handle DocBook dtd's.
--
Bob Bernstein http://www.ruptured-duck.com
Reply to: